Risk factors for acute myeloid leukemia and myelodysplastic syndromes

Kim Robien

Funding for this project will be included in the competing renewal application of the IWHS (to be submitted 3/2014). We will include $3,000 per cohort to fund extraction of covariate data. We will also include funding for a data manager at the University of Minnesota who will perform data collection, cleaning and harmonization for this project.

Myelodysplastic syndromes (MDS) are a group of clonal hematologic disorders that result in dysplastic and ineffective hematopoiesis. Individuals with MDS have a high risk of progressing to leukemia, with approximately 30% expected to develop AML. Incidence of MDS rises with age, with the majority of cases diagnosed after age 60 years. Survival of MDS patients is poor, with estimates below 50%. MDS became reportable to the SEER registry in 2001. _x000D__x000D_Little is known about the etiology of MDS, although chemotherapy or radiation therapy from a previous cancer are known risk factors. Recent analyses from two cohorts, the NIH-AARP cohort and the Million Women Study, have also implicated overweight and obesity as risk factors for MDS (Ma et al. AJE 2009; Murphy et al. BJC 2013). Obesity is also one of the few established risk factors for AML, suggesting that the two disorders share some overlapping etiology. This is not surprising given the high prevalence of AML in individuals with a history of MDS._x000D__x000D_Given the low incidence rate of MDS (3.0 per 100,000), the power to evaluate MDS in most cohort studies is quite low. For example, 75 MDS cases have accrued in the Iowa Women's Health Study since the disease became reportable in 2001. The Cohort Consortium provides an opportunity to overcome this limitation in sample size by conducting a pooled analysis. _x000D__x000D_Since approximately 30% of MDS cases will progress to AML, we will focus on the well-established risk factors for AML - cigarette smoking (increased risk), aspirin (decreased risk) and obesity (increased risk). In a preliminary analysis of risk factors for MDS in the IWHS, we observed an increased risk of MDS in women who reported being overweight (HR=2.35, 95% CI 1.34-4.17) or obese (HR=1.75, 95% CI 0.83-3.71) at baseline. In addition, we also observed suggestive associations with increased BMI at age 18 years and subsequent risk of MDS, suggesting that early adulthood weight may also influence risk. The timing of overweight and obesity has not been explored in previous studies and would be evaluated further in our proposed project._x000D__x000D_Previous case-control studies have reported associations between smoking and MDS (Strom et al. 2005; Nisse et al. Leukemia 1995). In preliminary analysis of IWHS data, risk estimates for smoking (both ever/never and pack years) were greater than 1 but non-significant with wide confidence intervals. Similarly, power was limited to evaluate aspirin and other NSAIDs in IWHS. This association has not been evaluated in previous studies and deserves further study given the protective effect observed in AML._x000D__x000D_Finally, little is known about the etiology of AML in individuals with a personal history of MDS. To date, the ability to predict which individuals will progress to AML is limited. In this project, we propose to conduct an exploratory analysis of anthropometric and lifestyle characteristics that predict which MDS patients will progress to AML.

The overall goals of this analysis are to evaluate lifestyle and anthropometric risk factors for MDS and to identify overlapping risk factors for MDS and AML.

In this cohort consortium project, we will evaluate the following specifics aims:_x000D_1) Identify lifestyle and anthropometric risk factors for MDS, with a specific focus on risk factors that predict AML._x000D_2) Compare risk factors for AML and MDS_x000D_3) Explore predictors of AML in women who were previously diagnosed with MDS.

Proportional hazards regression will be used to evaluate associations between lifestyle and anthropometric variables and risk of MDS or leukemia. Exposures and covariates will be defined using standard categories across study populations. BMI will be categorized using the World Health Organization Criteria. Smoking will be evaluated as current/former/never smoking, as well as pack years where available. Ever use of NSAIDs and other anti-inflammatory drugs will be used, as well as dose and duration when available.

MDS and AML following MDS are both rare; therefore, few prospective cohorts have sufficient power to evaluate the relationship between risk factors and these cancers.

50 cases

Diagnoses of interest:_x000D_AML (ICD-O-3 codes: 9840, 9861, 9866-9867, 9871-9874, 9891-9897, 9910, 9920)_x000D_MDS (ICD-O-3 codes:9980, 9982-9987, 9989)_x000D_Date of diagnosis_x000D_Method of cancer ascertainment_x000D_Date of death, date lost to follow up or study end date

NSAID use: (yes/no, dose and duration if available)_x000D_Smoking: (never/former/current, pack years)_x000D_Anthropometric data: BMI, WHR, height_x000D_Farm/rural residence_x000D_Physical activity_x000D_Previous history of cancer

Age at baseline or date of birth_x000D_Sex_x000D_Education_x000D_Race/ethnicity_x000D_Alcohol consumption

No